In the OpenCL spec it looks like ‘isequal()’ is only defined for half, double, and float. Why don’t I have this operator for integer types? I would’ve expected it to be the opposite, as it is unusual to compare floating points for equality.
I believe:
int4 comp, a, b;
comp = a == b;
will give you the equivalent.
I think the isequals() exists because it returns an int type for float arguments.